Investigating the coverage of the Arkansas All-Payer Claims Database for examining health disparities related to persistent poverty areas in colorectal cancer patients.

PURPOSE

We aimed to (1) determine the extent of coverage of colorectal cancer patients in Arkansas All-Payer Claims Database (APCD), (2) assess coverage difference between persistent poverty and other areas, and (3) identify patient, tumor, and area factors associated with inclusion in APCD.

METHODS

Data were from 2018 to 2020 Arkansas APCD linked with 2019 Arkansas Central Cancer Registry (ACCR). We constructed four cohorts to assess APCD's coverage of CRC patients: (Cohort 1) ≥ 1 day of medical coverage in APCD in 2019; (Cohort 2) APCD coverage in the diagnosis month; continuous APCD coverage in the 30; Year around diagnosis (six months before to five months after diagnosis month) (Cohort 3); or until death within six months (Cohort 4). We compared proportions in the cohorts by area persistent poverty designation. Logistic regressions identified factors associated with inclusion in APCD cohorts.

PATIENT SELECTION

CRC patients diagnosed in 2019 from ACCR, excluding in situ disease.

RESULTS

Of the 1,510 CRC patients diagnosed in 2019, 83% had ≥ 1 day of medical coverage in 2019 APCD (Cohort1), 81% had coverage in the diagnosis month (Cohort 2), and 63% had continuous coverage in the year around diagnosis (Cohort 3). Additionally, 11% died within six months but had continuous coverage until death (Cohort 4, 74%). No coverage difference was found between persist poverty and other areas. Age and primary payer type at diagnosis were the main predictors of inclusion in APCD.

CONCLUSION

Arkansas APCD had high coverage of Arkansas CRC patients. No selection bias by area of persistent poverty designation was present.